An Optimal Rule Based Energy Management System for a Hybrid Electric 2-Wheeler Based on Dynamic Programming Technique

Hybrid electric vehicles (HEV) are quite popular solution because of their low fuel consumption and extended range of driving compared to other solutions available. While the fuel consumption depends on the selection of hybrid architecture, the other critical parameter that affects the fuel consumption (Fuel Economy) is the energy management among the different hybrid components. After finalizing the hybrid vehicle architecture, different strategies can be adopted for energy management viz. rule based, fuzzy logic based, Dynamic Programming (DP), ECMS, and PMP etc. Each of these approaches possess some advantages and disadvantages. One of the simple approaches to develop energy management system (supervisory control) is ‘Rule Based’ strategy. However, the outcome of the rule-based strategies heavily depends on selection of optimal rules for strategy development. The other alternative for developing an optimal control strategy is Dynamic Programming (DP). Though DP is a very good candidate for providing optimal results, but due to very high computational requirements it may not be suitable for real time applications like hybrid electric vehicles (HEV). This paper examines the potential of using dynamic programming offline to derive optimal rules for ‘Rule Based’ strategy. In this paper, DP has been used on a standard driving cycle (WMTC) to generate rules to form a ‘Rule Based’ energy management strategy. A 2-wheeler hybrid electric vehicle is modeled using Simulink and Simscape in MATLAB. Finally, results obtained from the simulation of proposed ‘Rule Based’ strategy (with rules generated offline) with vehicle model on WMTC are compared with the results obtained using real time DP.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01696627
  • Record Type: Publication
  • Source Agency: SAE International
  • Report/Paper Numbers: 2019-26-0128
  • Files: TRIS, SAE
  • Created Date: Jan 15 2019 11:14AM